Share Email Print
cover

Proceedings Paper

Field crop extraction robust to illumination variations based on specularity learning
Author(s): Zhenghong Yu; Cuina Li; Huabing Zhou
Format Member Price Non-Member Price
PDF $14.40 $18.00

Paper Abstract

In this paper, we proposed an illumination-invariant crop extraction method based on specularity learning. Several useful contextual cues including object appearance and location inspired by recognition mechanism of human beings were introduced and integrated to machine learning architecture, generating a well-trained highlight region classifier. Combing with the Hue-intensity Look-up table and super-pixel techniques, the classifier gives the final extraction result. Comparing experiment confirmed the validity and feasibility of our method.

Paper Details

Date Published: 14 December 2015
PDF: 7 pages
Proc. SPIE 9815, MIPPR 2015: Remote Sensing Image Processing, Geographic Information Systems, and Other Applications, 98151U (14 December 2015); doi: 10.1117/12.2205851
Show Author Affiliations
Zhenghong Yu, Guangdong Polytechnic of Science and Technology (China)
Wuhan Institute of Technology (China)
Cuina Li, China Meteorological Administration (China)
Huabing Zhou, Wuhan Institute of Technology (China)


Published in SPIE Proceedings Vol. 9815:
MIPPR 2015: Remote Sensing Image Processing, Geographic Information Systems, and Other Applications
Jianguo Liu; Hong Sun, Editor(s)

© SPIE. Terms of Use
Back to Top